CSK6 Basic Version Visual Voice AI Development Kit Face Recognition ZEPHYR Development Board 16MB Norflash

CSK6 Basic Version Visual Voice AI Development Kit Face Recognition ZEPHYR Development Board 16MB Norflash

Description:
- This kit is based on the CSK6011A chip and is developed and designed for visual+voice artificial intelligence scenarios. It has rich hardware peripherals and complete speech and image functions, and is connected to large model services. It provides rich algorithms such as speech recognition, sitting posture detection, facial recognition, object classification, and supports flexible combination use.
- Support for autonomous adaptation of AI algorithm models: Equipped with AI training inference tool LNN tailored for CSK chips, Al algorithm can quickly and accurately match chips. Seamless integration of autonomous quantitative training and deployment inference, algorithm migration completed in 15 minutes.

Voice Recognition:
Off-line voice recognition:
- Misawakening: <= 1 times/72 hours
- Wake up rate: >=95%
- Recognition rate: >=93%
Big model dialogue:
- Universal interaction success rate: > 90%
- Intelligent user manual Q&A accuracy: > 95%
- Full link response time: < 1.5s

Visual Recognition:
- Sitting posture detection: Based on an intelligent algorithm with 11 bone points, it accurately identifies 4 types of bad sitting postures.
- Head and shoulder+gesture+facial recognition: supports human head and shoulder recognition at a distance of 1-3.5 meters; Accuracy of 95%, response speed of 24ms.
- Object recognition: Open source, supporting 90+object classifications.

Cognitive AI:
- Chinese English translation: supports mutual translation between Chinese and English; Millisecond translation speed; Bleu score > 25
- Word evaluation: supports English evaluation including accurate pronunciation and understanding of word meanings; Support phoneme level recognition; Evaluation accuracy > 98%; Identification response time < 500ms
- Chinese word segmentation: Using advanced neural and its segmentation theory and technology, supporting the segmentation of Chinese sentences

Hardware:
- Main control chip: onboard CSK6011A high performance AI chip, main frequency up to 300MHz
- Flash: onboard 16MB norflash
- Camera module: GC0328 DVP camera module, 30W pixel
- Speaker: support broadcast
- Microphone: 2-channel analog silicon microphone, support recording
- Touch screen: 320 x 240; screen hardware interface - SPI; touch hardware interface - I2C
- Network module: onboard ESP32-C3 network module, support WiFi & BLE function

Files: https://docs2.listenai.com
Videos: https://b23.tv/lbpiHZb

Package Included:
- 1 x Development Kit